CN101038593A - Method and apparatus for representing a group of images - Google Patents
Method and apparatus for representing a group of images Download PDFInfo
- Publication number
- CN101038593A CN101038593A CNA2007100852745A CN200710085274A CN101038593A CN 101038593 A CN101038593 A CN 101038593A CN A2007100852745 A CNA2007100852745 A CN A2007100852745A CN 200710085274 A CN200710085274 A CN 200710085274A CN 101038593 A CN101038593 A CN 101038593A
- Authority
- CN
- China
- Prior art keywords
- mass
- image
- tone
- descriptor
- value
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 30
- 238000009795 derivation Methods 0.000 claims description 5
- 238000013459 approach Methods 0.000 claims description 3
- 230000008878 coupling Effects 0.000 claims description 3
- 238000010168 coupling process Methods 0.000 claims description 3
- 238000005859 coupling reaction Methods 0.000 claims description 3
- 230000005055 memory storage Effects 0.000 claims 1
- 239000003086 colorant Substances 0.000 abstract 1
- 238000005516 engineering process Methods 0.000 description 13
- 230000006870 function Effects 0.000 description 13
- 238000010586 diagram Methods 0.000 description 9
- 230000006978 adaptation Effects 0.000 description 8
- 238000005259 measurement Methods 0.000 description 6
- 239000004744 fabric Substances 0.000 description 4
- 230000000007 visual effect Effects 0.000 description 4
- 239000000654 additive Substances 0.000 description 2
- 230000000996 additive effect Effects 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 238000012935 Averaging Methods 0.000 description 1
- 238000005162 X-ray Laue diffraction Methods 0.000 description 1
- 238000009825 accumulation Methods 0.000 description 1
- 238000005267 amalgamation Methods 0.000 description 1
- 230000001174 ascending effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000009432 framing Methods 0.000 description 1
- 230000008676 import Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/90—Determination of colour characteristics
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
- G06F16/58—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
- G06F16/583—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content
- G06F16/5838—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content using colour
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/70—Information retrieval; Database structures therefor; File system structures therefor of video data
- G06F16/73—Querying
- G06F16/732—Query formulation
- G06F16/7328—Query by example, e.g. a complete video frame or video sequence
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/70—Information retrieval; Database structures therefor; File system structures therefor of video data
- G06F16/78—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
- G06F16/783—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content
- G06F16/7847—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content using low-level visual features of the video content
- G06F16/785—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content using low-level visual features of the video content using colour or luminescence
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/56—Extraction of image or video features relating to colour
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Library & Information Science (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Computational Linguistics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
A method of representing a group of images comprises determining the values of one or more dominant colours for the group of images and deriving a dominant colour representation expressing the group of images in terms of one or more of said dominant colour values.
Description
The application is that application number is 200410063464.3, the applying date is on July 5th, 2004, denomination of invention is divided an application for the application for a patent for invention of " method and apparatus of presentation video group ".
Technical field
The present invention relates to the method and apparatus that a kind of presentation video group is especially used the color representation image sets and is used to search for and recover image.
Background technology
The technology that the color that has various known use visual properties for example to present in the image comes presentation video.For example, in a known technology, the designated color value of each pixel, and can obtain color histogram by frame (bin) being set and the pixel count in the image counted for a plurality of color value scopes, described pixel has a color value in each scope.Color histogram can be used for presentation video then.In another known technology, the one or more mass-tones in the image are identified, and this mass-tone is used to indicate image.
What the present invention was concerned about especially is image sets.For example, image sets can be video image (frame or picture) sequence, or image any image sets in related with it any source in some way.For example, image sets can be single scene or the cinestrip in the image.The term of such image sets well known in the prior art is GroupOfFrames/GroupOfPictures (frame group/image sets).Below this term will be known as GoFGoP.In this explanation, the term image will be used to the frame/picture in the explanation group, and no matter whether it is frame of video or camera lens or tableaux.In addition, term image and image-region are interchangeable, unless be conspicuous from context.
A method that is used for the presentation video group is to select a single image from image sets, and with the expression of this single image as the entire image group.Use the known technology of expression single image to represent this single image then.For example, this single image can be occur in the image sequence at first or last image, perhaps image sets is analyzed to discern the image in described group according to interested visual properties, this image is represented described image sets especially in some aspects.
Another method is for assembling image sets.Existing MPEG-7 Visual Standard (visual standard) (ISO/IEC 15938-3) allows to use GoFGoP color description symbol that the color in video-frequency band or the image sets is described.For example, this book of writing at Manjunath, Salembier and Sikora: in the part 13.5 of Introduction to MPEG-7 Multimedia content descriptioninterface (ISBN 0-471-48678-7) detailed introduction is arranged.Wherein introduced the technology of three kinds of focused image groups: average, intermediate value and common factor.
In each technology, color histogram is obtained each image of being used for organizing, as mentioned above.In the technology of averaging, color histogram is accumulated, and divided by producing an average histogram, wherein N is the quantity of image in described group to the frame of each accumulation (bin) value by N then.In the intermediate value technology, for each frame value, the histogram value of image sets is with ascending order or descending sort, and is that each frame is specified intermediate value.For each frame, obtain the histogram that occurs simultaneously by adopting the minimum histogram value in the image sets histogram.In other words, the common factor histogram is represented the quantity of the pixel of the particular color that occurs in all images or color gamut (being equivalent to a frame).
So, in each case, use a upgradeable color descriptor (seeing the part 13.4 in the above-mentioned book) to represent to assemble histogram (average, intermediate value and common factor), it is the value that the encoding scheme based on Haar transform is applied to color histogram.
The paper of Tong Lin and Hong-Jiang Zhang " carrying out the automatic video frequency scene by continuous pictures marshalling extracts " relates to continuous picture is organized into scene.A framing that forms cinestrip is analyzed to determine the mass-tone object in each frame, determined to continue the mass-tone object of entire frame group then.This will produce a mass-tone histogram for a cinestrip.Mass-tone histogram to different cinestrip compares.If the correlativity between two cinestrip is high, then described cinestrip is grouped into a scene.
Summary of the invention
Various aspects of the present invention are illustrated in accompanying Claim.
Description of drawings
With reference to the accompanying drawings embodiments of the invention are described, wherein:
Fig. 1 is the block scheme of system according to an embodiment of the invention;
Fig. 2 a, 2b and Fig. 3 to 5 merge the coordinate diagram of the method for image descriptor according to an embodiment of the invention for explanation;
Fig. 6 represents that the combined diagram picture is to form a super graph picture.
Embodiment
System is shown in Figure 1 according to an embodiment of the invention.This system comprises a control module 12, the computing machine that for example is used for the control system operation, this control module 12 comprises a storer and a processor at least, described system comprises that also one is connected to the display unit 14 of control module 12, for example monitor is used to show output and indication (pointing) equipment 16 of comprising image and text, for example mouse is used for to control module 12 input instructions.This system also comprises the image data base 18 of a plurality of image sets of a storage digital form and the descriptor database of memory descriptor information, and with more detailed explanation below, described descriptor information is used for being stored in every group of image of image data base 8.In this example, every group of image is corresponding to the cinestrip in the video sequence.There are various known technology to be used for video sequence is divided into cinestrip, will it be elaborated here.The present invention can be applied to any image sets, or image-region, comprises the zone in the image.Image can perhaps, for example, can have an identifier to indicate which image to belong to identical group with the form storage of group.
Each of image data base 18 and descriptor database 20 all is connected to control module 12.This system also comprises a search engine 22, and it is a computer program under control module 12 controls, and operates on the descriptor database 20.
In the present embodiment, the element of this system is provided on the single position, image library for example, and wherein the parts of this system are forever connected.
Each image has an associated image descriptor, and it is according to each image of main color showing in the image.In the present embodiment, image descriptor is basically as in our acquisition described in the pending application WO00/67203 simultaneously, and its content is by with reference to being merged in this paper.
The short-summary that obtains image descriptor is as follows.Each image has a plurality of pixels, and each pixel has associated color value, for example a RGB in relevant color space.
Color value and corresponding pixel are combined into the group so that determine mass-tone and which color are corresponding to each mass-tone in color gamut.This can use suitable clustering algorithm to realize, broad sense Laue moral algorithm for example, described in the part 13.3.1 in above-mentioned MPEG-7 book.
A bunch centre of moment that obtains from this cluster process is used as the mass-tone value, and the pixel groups that is used to form each bunch is stored for calculating other territory (weights and color variance), and is as described below.
Selectable, can use histogram method to obtain mass-tone, as described in the WO00/67203.
Under the sort of situation, the color value by selecting predetermined quantity in relevant color space or the scope of color value obtain the color histogram of image, and the pixel count that has each color value or have a value in the relevant range in the image is counted.
Usually, histogram will have one or more peak values, and each peak value (perhaps its subclass, for example peak-peak of predetermined quantity) is selected as mass-tone, simultaneously color value/pixel in color gamut about mass-tone by cluster.
In case color value by cluster, has then just been determined the color variance yields for each mass-tone, expression is that the center changes with the mass-tone for the color value of each each bunch of mass-tone.Mass-tone can be counted as the mean value of color distribution in relevant bunch.
Can use following formula to represent the calculating of variance:
Wherein j indicates color component, m
jBe j component of mass-tone, p
KjBe j component of k pixel value, and summation has surpassed and a mass-tone corresponding N pixel of being considered.
Described descriptor also comprises a weighted value for each mass-tone, and it is the measuring of relevant effective value of each mass-tone in the image.In this example, weighted value be corresponding to the mass-tone value bunch in pixel count and image in the ratio of sum of pixel.Weighted value can be expressed as percentage.
Mass-tone value and their each variance and weighted value are combined to form the color descriptor of image.Descriptor also can have other component, frequency n for example, the quantity of expression mass-tone.Descriptor also can comprise covariance value Cij, wherein for each mass-tone and bunch, and variance yields, i and j are represented the color component in the relevant color space.
The color descriptor of each image is stored in the descriptor database.
Notice that emphatically dominant color descriptors represents different with the histogram of image.Dominant color descriptors comprises the value of the mass-tone in the image, and it is determined in previous treatment step.The histogram of image represents to comprise mass-tone, but does not discern mass-tone.Dominant color descriptors also can comprise other value, and for example corresponding to the variance about the color distribution of this mass-tone or each mass-tone, but histogram is represented not relate to and calculated or determine this variance or other value.For example, other component of dominant color descriptors can comprise the weighted value of the influence of the mass-tone in the presentation video, the mass-tone number in the image and corresponding to the space homogeney of the pixel of the mass-tone in the image.
According to one first embodiment, the following acquisition of descriptor of set of diagrams picture group.
The image descriptor of each image in the image sets is resumed from descriptor database.Image descriptor is combined to form a group descriptor then.
Group descriptor has the form that is similar to image descriptor (mass-tone, variance, weighted value etc.).Preferably, the mass-tone number in the group descriptor is between 1 and 8.Yet the mass-tone number can be unrestricted or a predetermined maximum can be set.Mass-tone number in the image descriptor also can be unrestricted or can be limited by a predetermined maximum value.The maximal value of image descriptor can be identical with the maximal value of group descriptor, makes that for example, image descriptor has the mass-tone of Duoing than group descriptor.
In this example, two images are arranged, and two image descriptors separately are by following merging in one group.
Summarize, image descriptor based in the color space bunch the degree of approach by in the combined diagram picture bunch and combined.
Fig. 2 a and 2b be in the color space of two images bunch abstract representation.Fig. 2 a represents first image, and Fig. 2 b represents second image.In each case, circle is represented bunch of mass-tone separately.For the purpose of simplifying, color space illustrates with two dimensional form, though color space is in the three-dimensional usually.In addition, bunch needn't be, but illustrated like this for the purpose of simplifying corresponding to the circle in the color space, and any indication of the weighted value that do not provide bunch of circle.Below, such as being described to a bunch descriptor in the expression shown in Fig. 2 a and the 2b expression of (just, according to mass-tone and each bunch).
Two bunches of descriptors of Fig. 2 a and 2b are combined to form bunch super descriptor.This is shown in Figure 3.
Then, use the suitable range observation in the color space, the distance between every pair bunch (as defined above) in the super descriptor is determined.In this example, described distance is the Euclidean distance in the color space.
In this example, two mass-tones and two bunches are arranged in first image, and in second image, have three mass-tones and three bunches, respectively from 1 to 5 numbering.Each mass-tone is represented by the fork among Fig. 2 to 5 corresponding to a point in the RGB color space.Distance between a pair of bunch is the distance between bunch centre of moment in the 3-D rgb space.Distance between every pair of mass-tone in first and second images is calculated, comprise in the present example from identical image bunch.The mass-tone that selection provides the minor increment measurement is right.
Below, merge corresponding to bunch quilt of these two mass-tones.
In this example, as shown in Figure 3, two hithermost bunch is bunches 1 and 3, and they are by following merging.
Main or the representational color value that merges bunch is the weighted mean of two bunches mass-tone, and wherein weighted value is as above defined.Therefore, for two mass-tone m
1, m
2Weighted value W1 separately, W2, the mass-tone m of described merging has value:
m=w
1m
1+w
2m
2
W wherein
1, w
2Be respectively relevant weighted value, W1/W1+W2 and W2/W1+W2.
Two bunches the variance of use to merge, merging bunch variance also calculated.In this example, each color component is independently handled, and supposition merge bunch variance be the weighted sum of two Gaussian distribution.The formula of the variance of this merging below will producing bunch:
σ wherein
1 2, σ
2 2Be the variance of sub-clustering, m
1, m
2Be their average, and w
1, w
2Be relevant weighted value, as defined above.
The weighted value W that merges bunch is W1+W2.
Merge and bunch to be counted as new bunch, have aforesaid weighted value, mean value and variance.
This is illustrated among Fig. 4, and wherein bunch 1 and 3 quilts merge to form one new bunch 6.
Then carry out another time iteration of described combining step.Two bunches 1 and 3 that merge in the iteration in the first time are excluded outside further considering, and by bunches 6 replacing of merging, as shown in Figure 4.
Then by identification in color space comprise merge bunches 6 nearest bunch to and merge them, combining step is repeated, and outlines as top.
In the present example, in second time iteration, bunches 2 and 4 be nearest bunch right.They are merged to produce one new bunch 7, and it has mass-tone, weighted value and variance, and they obtain as mass-tone, weighted value and the variance by above-mentioned Fig. 2 and 4.What merge bunches 7 replaces bunches 2 and 4, as shown in Figure 5.
Merge iteration and be repeated until that predetermined condition is satisfied.For example, predetermined condition can be merge by continue up to stay predetermined quantity and bunch (merging bunch and, in first and second images original reservation bunch).Selectable, predetermined condition can be merge by continue up to each reservation bunch between distance greater than a set-point.Selectable, this method can comprise the iteration of predetermined quantity.More than one predetermined condition can be combined.
In the present circumstance, merge be repeated until stay predetermined quantity (three) bunch.
In the above example, merging bunch further is being considered in the iteration.Yet, also they can be got rid of outside further iteration.In addition, in the above example, first and/or any subsequently iteration in, bunch can with other bunch merging in the identical image, but selectable, bunch can be only and bunch merging that appears in the other image.This has reduced the number of range observation.
Though according to bunch being illustrated, should be appreciated that merging works to the value in the descriptor, just mass-tone, variance and weighted value, and do not need analytic manifold itself.
In case finished iteration, what stay bunch just is used to form group descriptor.Especially, for each last bunch, there are representational or main color, variance separately and a weighted value separately.They are combined to form group descriptor together with the number of times m of the last number of clusters of expression.Group descriptor also can comprise other factor, for example index of color space, or the color quantizing that uses in the expression.This group descriptor is the GoFGoP dominant color descriptors.
In the above example, in group of picture, have only two images.Yet this method also can be applied to comprise the group more than two images.Bunch descriptor of each image in the group can be all combined, to form a super descriptor, as mentioned above.Selectable, image sets can be combined into the subimage group, two or three groups for example, and in sequence or not in sequence, the descriptor combination in a similar manner of son group then.
When the huge amount of image in the group, top method is potential relate to enormous quantity bunch, and relate to enormous quantity bunch between the calculating of distance.
Consider above-mentionedly, the variance of said method has been considered such fact: most images that video cinestrip or video are concentrated will be that very similar and corresponding descriptor will be similar.This means that most bunches can be merged and do not have heavy losses of precision on the basis of every frame.
What in more detail, variance was considered is the image of set of diagrams picture in the sequence.As in above-mentioned example, image descriptor is derived and is stored in the descriptor database.Bunch descriptor of first and second images in the set of diagrams picture is resumed.Then, the distance between each of image pair bunch is determined.If any bunch between distance less than predetermined threshold value, then bunch to being merged.The merging of image pair bunch and any remaining bunch be concentrated into a super descriptor, as described in the above example.Then by merging close bunch and form a new super descriptor, bunch descriptor of next image is resumed and merges with the super descriptor of preceding two images in an identical manner.When bunch descriptor of all images all had been considered, the super descriptor that uses all remaining bunch amalgamation results to obtain was as described in first example.
In the superincumbent explanation, bunch descriptor is derived.Selectable, image can be resumed or provide, and just can derive descriptor from image before carrying out bunch merging.
In the first above-mentioned embodiment, derive the GoFGoP descriptor by the descriptor of each image in the assembly group.
In a second embodiment, image is assembled in image or pixel domain, derives dominant color descriptors to produce the GoFGoP dominant color descriptors from the image of assembling then.Therefore, if N image 8 arranged in group, each image comprises m * m pixel, then super graph as 9 can be counted as (pel array of N * m) * m, as shown in Figure 6.
Form contrast with first embodiment, second embodiment does not use the image descriptor of every width of cloth image, but directly image is worked.
After image is assembled, use the above-mentioned technology relevant with first embodiment, derive dominant color descriptors from the super graph picture.
The advantage of second embodiment is: do not have loss of significance in extracting processing.Yet, when considering all images, just need a large amount of complicacy, especially need a large amount of storeies.In order to overcome the problem of complicacy, on time and/or space, can carry out secondary sample to image.
Above, image is assembled in image or pixel domain, derives dominant color descriptors to produce the GoFGoP dominant color descriptors from the image of assembling then.Selectable, each image is assembled the histogrammic form of each image (for example, with) and is derived dominant color descriptors from the histogram of assembling in color gamut.For example, can from storer, derive or recover the color histogram of each image.Then, the color combining histogram is to form a super image histogram by they are added up, and it can be by forming the picture number standardization of super graph picture.At last, use the above-mentioned technology relevant to derive dominant color descriptors from super image histogram with first embodiment.In other words, super graph is selected as histogrammic peak value (mass-tone) and variance separately and weighted value.
Can use other technology of before deriving dominant color descriptors, assembling image.For example, not that the histogram of every width of cloth image is added up, but the average of computed image group, intermediate value or common factor.Derive the mass-tone group descriptor from the average, intermediate value or the common factor histogram that obtain then.
In the variance of second embodiment, it is used for assembling in image/pixel territory and color/histogram territory, and image sets is in time by secondary sample, and is as described below.In this example, bunch descriptor of every width of cloth image is derived or is recovered.Which width of cloth image similarity decision based on bunch descriptor uses derive the GoFGoP descriptor.
Form initial super graph picture from first image in the sequence of image sets.Following map is abandoned satisfies predetermined decision condition up to an image and the similarity that is added between bunch descriptor of last image of super graph picture (being initially first image).If an image satisfies this condition, then it just is added to the super graph picture.Following map in will organizing then compares up to predetermined condition with the last image that is added to the super graph picture and is satisfied once more, and perhaps all images in group all is considered, or the like.Derive dominant color descriptors from the super graph picture that obtains then.
One of image similarity may determine that criterion is the value of the adaptation function between each dominant color descriptors, for example described in we the common pending application WO00/67203, perhaps uses as at the adaptation function described in the above-mentioned MPEG-7 book.Other criterion can be the result that " immediately " recited above merges.If all bunches all have been merged in existing descriptor, then " decision criterion " in this case will be satisfied.Another method of avoiding extracting the mass-tone of all images is to calculate a coarse color histogram and use this histogram adaptation function value as criterion.
All these criterions all need to specify other parameter: under first kind of situation, are lower than its adaptation function value and are considered to little threshold value, and the merging threshold value under second kind of situation.A selectable method is to adapt to this threshold value so that the picture number of collecting can not surpass the boundary of a regulation, and this method is specially adapted to the situation of finite memory.
The 3rd embodiment derives the GoFGoP dominant color descriptors from the dominant color descriptors group of image sets.More special, for the set of diagrams picture, from each dominant color descriptors of descriptive data library searching (, then deriving them) if perhaps they also do not derive.
For each dominant color descriptors, the distance between each in the residue descriptor in described descriptor and this group is measured.This will produce one group of range measurements for each descriptor, and described range measurements can be thought that each descriptor provides a total range measurements by adding up.Descriptor is represented in the selected conduct of descriptor with the total range measurements of minimum, and is counted as the GoFGoP dominant color descriptors.
Select the additive method of representational descriptor to be used, preferably comprise test or comparison, it relates to some image descriptors of the image of this group at least.As other example, be used for whole descriptor if be implemented, then this can be based on finishing as the distortion measurement that defines in MPEG-7.
Can abandon very near descriptor in advance calculates to reduce.
Illustrated above and be used to derive the GoFGoP descriptor, especially the whole bag of tricks of GoFGoP dominant color descriptors.
For the GoFGoP descriptor various application are arranged, for example be used for search and recover image sets.For example, the user wishes to search for the image sets corresponding with input picture or image sets.
The explanation of following summary a kind of searching method.
With reference to Fig. 1, use a proper device by the user, for example scanner or digital camera perhaps by selecting an inquiry image the image range that shows from computing machine, have perhaps been imported an inquiry image by a zone selecting any this image.The dominant color descriptors of this image is derived, as mentioned above.To inquire that then dominant color descriptors and each the GoFGoP dominant color descriptors that is stored in the descriptor database compare.The GoFGoP dominant color descriptors is identical with the form of single image dominant color descriptors, so mate, for example, uses as the adaptation function in WO00/67203 or analog, perhaps as at the adaptation function described in the part 13.3.2 of above-mentioned MPEG-7 book.The single image descriptor with being stored in the described database that the inquiry descriptor also can be selected compares.
The result of adaptation function is organized, and adaptation function is indicated the image sets of immediate coupling for it and is resumed.One or more images of immediate coupling are shown.
Can use the additive method that proposes inquiry.Can be by selecting the set of diagrams picture and extract aforesaid GoFGoP descriptor to propose inquiry for this group.For example,, for example, then can clearly select this group, use suitable algorithm to derive one " cinestrip " that comprises this key frame then by in a video, selecting a key frame by selecting a frame scope or implicity.
For example, can be provided in the image library according to system of the present invention.Selectable, database can be away from the control module of system location, can be by interim circuit telephone wire or for example be connected to control module in the internet by network for example.For example, image and descriptor database can be provided in the permanent storage device or on the portable data storage medium, for example CD-ROM or DVD.
As top be the form of computer system according to the described system of first embodiment of the invention.Computing machine can be a standard computer, uses suitable program described computing machine to be carried out programming to be used to carry out the method according to the embodiment of the invention.Described program can be stored on any suitable storage medium, comprises fixing or permanent reservoir or mobile storage means.Use specific hardware and/or software, comprise, for example certain chip can be made amendment to this system.Also can realize the present invention in specific suitable device, described device comprises specific hardware and/or software.
In the superincumbent explanation, color showing is illustrated according to the red, green and blue color component.Certainly, can use other expression, for example use tone, saturation degree and brightness to represent or the YUV coordinate system, the subclass of the color component in what its color space perhaps in office is for example only used the color harmony saturation degree in HIS.
Aforesaid embodiments of the invention have used the descriptor as image and image sets derivation.Yet image descriptor can be used for the zone of image, and similar GoFGoP descriptor can produce based on the zone of image.The zone can be a rectangle frame, perhaps can use the zone of difformity and size.Selectable, can be zone derivation descriptor, for example automobile, house or people corresponding to the image of target.Under any therein situation, can be all images or its part derivation descriptor.In addition, can be single image by a plurality of zones in the image that top method is applied to form set of diagrams picture zone and derive the GoFGoP descriptor.
In retrieving, the user for example uses indicating equipment by rotate a district describing image around it, rather than import the solid color inquiry or select an image block, so control module is derived a descriptor and is used it to retrieve in above-mentioned similar mode for this zone.In addition, not to utilize the image that is stored in the image data base to cause retrieval, and for example be to use image reading apparatus or digital camera that image is input in the system.Retrieving under such situation, this system can determine to come the zone for this image or image to derive descriptor again at first automatically or by the user.
Can use hardware or software to realize suitable aspect of the present invention.
In the above-described embodiments, use Gaussian function to come bunch distribution of each representative color is similar to, and these average values of a function, variance and covariance are used to descriptor value.Yet, can use other function or parameter to come component distributed and be similar to, for example, use the basis function of similar sine and cosine, descriptor is based on these functions.
Claims (7)
1. the method for a presentation video group (8), this method comprises:
For this image sets is determined one or more mass-tone values (1-5) and is represented according to the mass-tone of this image sets of one of one or more derivation in described mass-tone value performance, wherein some images in this image sets are to represent representative by the mass-tone according to one or more mass-tone values of this image separately at least, and wherein each mass-tone represents to have one or more components that comprise at least one mass-tone value, this method also comprises in conjunction with a plurality of described mass-tones and represents or select a mass-tone to represent representative as this group, it is characterized in that in color space the degree of approach based on mass-tone is in conjunction with the one or more components in the described expression.
2. the method for claim 1, wherein each mass-tone represents to have one or more components that are used for each mass-tone value alternatively that comprise, show the variance of image about the color distribution variance of mass-tone value, the weighted value that shows the influence of mass-tone in the image, the quantity of mass-tone and corresponding to the space homogeney of the pixel of the mass-tone in the image in the image.
3. the described method of any one claim as described above is included on time or the space described image sets (8) is carried out secondary sample.
4. method as claimed in claim 3 is wherein omitted according to the similarity of other image in they and described group or is comprised image.
5. as claim 3 or the described method of claim 4, wherein said secondary sample comprises the described step of claim 1.
6. the device of a presentation video group (8), this device comprises:
For this image sets is determined one or more mass-tone values (1-5) and the device of representing according to the mass-tone of this image sets of a performance of the one or more derivation in the described mass-tone value, wherein some images in this image sets are to represent representative by the mass-tone according to one or more mass-tone values of this image separately at least, and wherein each mass-tone represents to have one or more components that comprise at least one mass-tone value; And
Coupling apparatus is used for representing or selecting a mass-tone to represent representative as this group in conjunction with a plurality of described mass-tones, and in color space based on the degree of approach of mass-tone in conjunction with the one or more components in the described expression.
As claim 6 fast device, comprise the memory storage that is used for storing image data and/or descriptive data, treating apparatus and comprising is used for the control device that enforcement of rights requires the instruction of any one described method of 1 to 5.
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
EP03254263A EP1494132B1 (en) | 2003-07-04 | 2003-07-04 | Method and apparatus for representing a group of images |
EP03254263.1 | 2003-07-04 |
Related Parent Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CNB2004100634643A Division CN1311388C (en) | 2003-07-04 | 2004-07-05 | Method and apparatus for representing a group of images |
Publications (2)
Publication Number | Publication Date |
---|---|
CN101038593A true CN101038593A (en) | 2007-09-19 |
CN101038593B CN101038593B (en) | 2010-06-23 |
Family
ID=33427249
Family Applications (6)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CNA2007100852783A Pending CN101013446A (en) | 2003-07-04 | 2004-07-05 | Image group representation method and device |
CN2007100852745A Expired - Lifetime CN101038593B (en) | 2003-07-04 | 2004-07-05 | Method and apparatus for representing a group of images |
CN2007100852730A Expired - Lifetime CN101038592B (en) | 2003-07-04 | 2004-07-05 | Method and apparatus for representing a group of images |
CN2007100852779A Expired - Lifetime CN101017507B (en) | 2003-07-04 | 2004-07-05 | Image group representation method and device |
CNB2007100852764A Expired - Lifetime CN100501742C (en) | 2003-07-04 | 2004-07-05 | Image group representation method and device |
CNB2004100634643A Expired - Lifetime CN1311388C (en) | 2003-07-04 | 2004-07-05 | Method and apparatus for representing a group of images |
Family Applications Before (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CNA2007100852783A Pending CN101013446A (en) | 2003-07-04 | 2004-07-05 | Image group representation method and device |
Family Applications After (4)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN2007100852730A Expired - Lifetime CN101038592B (en) | 2003-07-04 | 2004-07-05 | Method and apparatus for representing a group of images |
CN2007100852779A Expired - Lifetime CN101017507B (en) | 2003-07-04 | 2004-07-05 | Image group representation method and device |
CNB2007100852764A Expired - Lifetime CN100501742C (en) | 2003-07-04 | 2004-07-05 | Image group representation method and device |
CNB2004100634643A Expired - Lifetime CN1311388C (en) | 2003-07-04 | 2004-07-05 | Method and apparatus for representing a group of images |
Country Status (5)
Country | Link |
---|---|
US (6) | US7630545B2 (en) |
EP (6) | EP2164044B1 (en) |
JP (2) | JP4553650B2 (en) |
CN (6) | CN101013446A (en) |
DE (2) | DE60330720D1 (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101038592B (en) * | 2003-07-04 | 2010-09-01 | 三菱电机株式会社 | Method and apparatus for representing a group of images |
Families Citing this family (44)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7813552B2 (en) * | 2004-09-23 | 2010-10-12 | Mitsubishi Denki Kabushiki Kaisha | Methods of representing and analysing images |
JP2006119723A (en) * | 2004-10-19 | 2006-05-11 | Canon Inc | Device and method for image processing |
JP4881034B2 (en) * | 2005-02-28 | 2012-02-22 | 富士フイルム株式会社 | Electronic album editing system, electronic album editing method, and electronic album editing program |
DE602006002242D1 (en) * | 2006-02-15 | 2008-09-25 | Jaycrypto Ltd | Method and device for generating initial values for a random number generator |
US20070201743A1 (en) * | 2006-02-28 | 2007-08-30 | Sharp Laboratories Of America, Inc. | Methods and systems for identifying characteristics in a digital image |
US8116573B2 (en) * | 2006-03-01 | 2012-02-14 | Fujifilm Corporation | Category weight setting apparatus and method, image weight setting apparatus and method, category abnormality setting apparatus and method, and programs therefor |
US20080002855A1 (en) * | 2006-07-03 | 2008-01-03 | Barinder Singh Rai | Recognizing An Unidentified Object Using Average Frame Color |
US7809185B2 (en) * | 2006-09-21 | 2010-10-05 | Microsoft Corporation | Extracting dominant colors from images using classification techniques |
ATE495501T1 (en) * | 2006-10-19 | 2011-01-15 | Koninkl Philips Electronics Nv | DESCRIPTORS FOR DOMINANT COLORS |
US7912279B2 (en) * | 2006-10-26 | 2011-03-22 | Qualcomm Incorporated | Automatic white balance statistics collection |
JP4895878B2 (en) * | 2007-03-19 | 2012-03-14 | 富士フイルム株式会社 | Representative color extraction method and representative color extraction apparatus |
CN101625753B (en) * | 2008-07-10 | 2012-11-21 | 辉达公司 | Grating for processing graph and rasterizing method |
US8401342B2 (en) * | 2009-01-16 | 2013-03-19 | A9.Com, Inc. | System and method to match images using topologically equivalent correspondences |
GB2461641A (en) * | 2009-07-08 | 2010-01-13 | Dan Atsmon | Object search and navigation |
JP5081948B2 (en) * | 2010-05-14 | 2012-11-28 | 株式会社バッファロー | Wireless relay device |
US8612441B2 (en) * | 2011-02-04 | 2013-12-17 | Kodak Alaris Inc. | Identifying particular images from a collection |
US9436704B2 (en) | 2012-02-07 | 2016-09-06 | Zencolor Corporation | System for normalizing, codifying and categorizing color-based product and data based on a universal digital color language |
US10460475B2 (en) | 2012-02-07 | 2019-10-29 | Zencolor Global, Llc | Normalization of color from a captured image into a universal digital color system for specification and matching |
US8600153B2 (en) | 2012-02-07 | 2013-12-03 | Zencolor Corporation | System and method for normalization and codification of colors for dynamic analysis |
WO2013184804A1 (en) * | 2012-06-06 | 2013-12-12 | Zencolor Corporation | System and method for normalization and codificaton of colors for dyanamic analysis |
US9607404B2 (en) | 2012-02-07 | 2017-03-28 | Zencolor Corporation | System for normalizing, codifying and categorizing color-based product and data based on a universal digital color system |
US9047633B2 (en) | 2012-02-07 | 2015-06-02 | Zencolor Corporation | System and method for identifying, searching and matching products based on color |
US9076076B1 (en) * | 2012-03-30 | 2015-07-07 | Google Inc. | Image similarity determination |
US10140537B2 (en) * | 2012-10-26 | 2018-11-27 | Daon Holdings Limited | Methods and systems for capturing biometric data |
US9147275B1 (en) | 2012-11-19 | 2015-09-29 | A9.Com, Inc. | Approaches to text editing |
US9043349B1 (en) | 2012-11-29 | 2015-05-26 | A9.Com, Inc. | Image-based character recognition |
US8837867B2 (en) * | 2012-12-07 | 2014-09-16 | Realnetworks, Inc. | Method and system to detect and select best photographs |
US9342930B1 (en) | 2013-01-25 | 2016-05-17 | A9.Com, Inc. | Information aggregation for recognized locations |
US20150371107A1 (en) * | 2014-06-23 | 2015-12-24 | Canon Kabushiki Kaisha | Material classification using brdf feature vector |
US9087357B2 (en) | 2013-10-16 | 2015-07-21 | Zencolor Corporation | System for normalizing, codifying and categorizing color-based product and data based on a universal digital color language |
US9424598B1 (en) | 2013-12-02 | 2016-08-23 | A9.Com, Inc. | Visual search in a controlled shopping environment |
US9477748B2 (en) * | 2013-12-20 | 2016-10-25 | Adobe Systems Incorporated | Filter selection in search environments |
US9536161B1 (en) | 2014-06-17 | 2017-01-03 | Amazon Technologies, Inc. | Visual and audio recognition for scene change events |
US10019143B1 (en) * | 2014-09-29 | 2018-07-10 | Amazon Technologies, Inc. | Determining a principal image from user interaction |
CN106204690B (en) * | 2015-05-04 | 2020-04-17 | 阿里巴巴集团控股有限公司 | Image processing method and device |
CN104951562B (en) * | 2015-07-08 | 2018-06-05 | 天云融创数据科技(北京)有限公司 | A kind of image search method based on VLAD dual adaptions |
US11256923B2 (en) | 2016-05-12 | 2022-02-22 | Arris Enterprises Llc | Detecting sentinel frames in video delivery using a pattern analysis |
US10097865B2 (en) | 2016-05-12 | 2018-10-09 | Arris Enterprises Llc | Generating synthetic frame features for sentinel frame matching |
US10388250B2 (en) * | 2017-04-18 | 2019-08-20 | Google Llc | Methods, systems, and media for modifying user interface colors in connection with the presentation of a video |
US10534973B2 (en) * | 2017-04-18 | 2020-01-14 | Google Llc | Methods, systems, and media for color palette extraction for video content items |
CN107832359B (en) * | 2017-10-24 | 2021-06-08 | 杭州群核信息技术有限公司 | Picture retrieval method and system |
US11682142B2 (en) * | 2019-12-20 | 2023-06-20 | Raytheon Company | Information weighted rendering of 3D point set |
US11657541B2 (en) * | 2020-08-07 | 2023-05-23 | Texas Instruments Incorporated | Parent-child cluster compression |
CN112887694B (en) * | 2021-01-26 | 2023-03-10 | 腾讯音乐娱乐科技(深圳)有限公司 | Video playing method, device and equipment and readable storage medium |
Family Cites Families (35)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US3786283A (en) * | 1970-12-29 | 1974-01-15 | Sansui Electric Co | Delay circuit for providing a variable delay characteristic |
US5051629A (en) * | 1988-06-20 | 1991-09-24 | Hopkins Michael C | Signal delay and reproduction device and method |
US5119186A (en) * | 1989-09-28 | 1992-06-02 | International Business Machines Corporation | Color mapping system and method |
US5164711A (en) * | 1990-02-08 | 1992-11-17 | International Business Machines Corporation | System and method for generating graphical output |
US5146192A (en) * | 1990-07-06 | 1992-09-08 | Hitachi Medical Corporation | Delay circuit of ultrasonic diagnostic apparatus using delay line comprising variable capacitance diode and inductor |
EP0542361B1 (en) * | 1991-11-14 | 1998-09-16 | Agfa-Gevaert N.V. | Method and apparatus for constructing a histogram in digital image processing by statistical pixel sampling |
US5318034A (en) * | 1991-11-27 | 1994-06-07 | Hitachi Medical Corp. | Electrically adjustable delay circuit and ultrasonic diagnosis apparatus |
JP2721107B2 (en) * | 1993-02-23 | 1998-03-04 | 日本電信電話株式会社 | Video feature processing method |
JP3419415B2 (en) * | 1993-07-07 | 2003-06-23 | 日本電信電話株式会社 | Video feature processing method |
US5802214A (en) * | 1995-12-08 | 1998-09-01 | Xerox Corporation | Method for determining and loading an image-dependent look-up table for generating an enhanced image representation |
US6075875A (en) * | 1996-09-30 | 2000-06-13 | Microsoft Corporation | Segmentation of image features using hierarchical analysis of multi-valued image data and weighted averaging of segmentation results |
JP3747589B2 (en) * | 1997-09-17 | 2006-02-22 | コニカミノルタビジネステクノロジーズ株式会社 | Image feature amount comparison device and recording medium storing image feature amount comparison program |
US6430312B1 (en) * | 1997-12-29 | 2002-08-06 | Cornell Research Foundation, Inc. | Image subregion querying using color correlograms |
JPH11259507A (en) * | 1998-03-12 | 1999-09-24 | Nippon Telegr & Teleph Corp <Ntt> | Video retrieval method and medium for storing video retrieval program |
US6285995B1 (en) | 1998-06-22 | 2001-09-04 | U.S. Philips Corporation | Image retrieval system using a query image |
US6345274B1 (en) * | 1998-06-29 | 2002-02-05 | Eastman Kodak Company | Method and computer program product for subjective image content similarity-based retrieval |
JP2000069297A (en) * | 1998-08-18 | 2000-03-03 | Hajime Matsuoka | Method for objectively evaluating image quality of irreversibly compressed image |
US6516100B1 (en) * | 1998-10-29 | 2003-02-04 | Sharp Laboratories Of America, Inc. | Method for image characterization using color and texture statistics with embedded spatial information |
AU1930700A (en) * | 1998-12-04 | 2000-06-26 | Interval Research Corporation | Background estimation and segmentation based on range and color |
US6272250B1 (en) * | 1999-01-20 | 2001-08-07 | University Of Washington | Color clustering for scene change detection and object tracking in video sequences |
AU753599B2 (en) * | 1999-02-05 | 2002-10-24 | Regents Of The University Of California, The | Color image processing method and apparatus thereof |
US6477269B1 (en) * | 1999-04-20 | 2002-11-05 | Microsoft Corporation | Method and system for searching for images based on color and shape of a selected image |
GB2349460B (en) * | 1999-04-29 | 2002-11-27 | Mitsubishi Electric Inf Tech | Method of representing colour images |
JP2001043371A (en) * | 1999-07-29 | 2001-02-16 | Canon Inc | Method and device for processing image and storage medium |
US6577759B1 (en) | 1999-08-17 | 2003-06-10 | Koninklijke Philips Electronics N.V. | System and method for performing region-based image retrieval using color-based segmentation |
KR20010081894A (en) * | 2000-02-18 | 2001-08-29 | 구자홍 | Multimedia Query System And Histogram Converting System Based On Contents |
KR100439371B1 (en) * | 2000-09-08 | 2004-07-09 | 엘지전자 주식회사 | Multimedia searching method using histogram |
US6788813B2 (en) * | 2000-10-27 | 2004-09-07 | Sony Corporation | System and method for effectively performing a white balance operation |
US7062084B2 (en) * | 2000-12-01 | 2006-06-13 | Sharp Laboratories Of America, Inc. | Method for image description using color and local spatial information |
US6944334B2 (en) * | 2001-02-01 | 2005-09-13 | Eastman Kodak Company | Color cross reference selection system and method |
DE60235477D1 (en) * | 2001-07-31 | 2010-04-08 | Aevora Beauty Concepts Inc | ADVANCED COSMETIC COLOR ANALYSIS SYSTEM |
EP1302865A1 (en) * | 2001-10-10 | 2003-04-16 | Mitsubishi Electric Information Technology Centre Europe B.V. | Method and apparatus for searching for and retrieving colour images |
US7415153B2 (en) * | 2002-09-20 | 2008-08-19 | International Business Machines Corporation | Color naming, color categorization and describing color composition of images |
EP2164044B1 (en) * | 2003-07-04 | 2012-04-18 | Mitsubishi Electric Information Technology Centre Europe B.V. | Method and apparatus for representing a group of images |
US20060291750A1 (en) * | 2004-12-16 | 2006-12-28 | Peyman Milanfar | Dynamic reconstruction of high resolution video from low-resolution color-filtered video (video-to-video super-resolution) |
-
2003
- 2003-07-04 EP EP09012972A patent/EP2164044B1/en not_active Expired - Lifetime
- 2003-07-04 EP EP10179677A patent/EP2302590A1/en not_active Withdrawn
- 2003-07-04 EP EP03254263A patent/EP1494132B1/en not_active Expired - Lifetime
- 2003-07-04 EP EP10179664A patent/EP2293250B1/en not_active Expired - Lifetime
- 2003-07-04 EP EP07018181A patent/EP1881455B1/en not_active Expired - Lifetime
- 2003-07-04 DE DE60330720T patent/DE60330720D1/en not_active Expired - Lifetime
- 2003-07-04 DE DE60317053T patent/DE60317053T2/en not_active Expired - Lifetime
- 2003-07-04 EP EP10181847A patent/EP2273451B1/en not_active Expired - Lifetime
-
2004
- 2004-06-25 US US10/875,237 patent/US7630545B2/en active Active
- 2004-07-05 CN CNA2007100852783A patent/CN101013446A/en active Pending
- 2004-07-05 CN CN2007100852745A patent/CN101038593B/en not_active Expired - Lifetime
- 2004-07-05 CN CN2007100852730A patent/CN101038592B/en not_active Expired - Lifetime
- 2004-07-05 JP JP2004198403A patent/JP4553650B2/en not_active Expired - Lifetime
- 2004-07-05 CN CN2007100852779A patent/CN101017507B/en not_active Expired - Lifetime
- 2004-07-05 CN CNB2007100852764A patent/CN100501742C/en not_active Expired - Lifetime
- 2004-07-05 CN CNB2004100634643A patent/CN1311388C/en not_active Expired - Lifetime
-
2007
- 2007-10-30 US US11/929,207 patent/US7676085B2/en not_active Expired - Lifetime
- 2007-10-30 US US11/929,139 patent/US7689044B2/en not_active Expired - Lifetime
- 2007-10-30 US US11/929,197 patent/US7715621B2/en not_active Expired - Lifetime
- 2007-10-30 US US11/929,220 patent/US7826661B2/en not_active Expired - Lifetime
- 2007-10-30 US US11/929,189 patent/US7684616B2/en active Active
-
2010
- 2010-06-08 JP JP2010131092A patent/JP4990383B2/en not_active Expired - Lifetime
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101038592B (en) * | 2003-07-04 | 2010-09-01 | 三菱电机株式会社 | Method and apparatus for representing a group of images |
Also Published As
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN1311388C (en) | Method and apparatus for representing a group of images | |
US11449540B1 (en) | Media fingerprinting and identification system | |
Chen et al. | Automatic detection of object-based forgery in advanced video | |
US6819797B1 (en) | Method and apparatus for classifying and querying temporal and spatial information in video | |
US6253201B1 (en) | Scalable solution for image retrieval | |
US20070030396A1 (en) | Method and apparatus for generating a panorama from a sequence of video frames | |
US20100049711A1 (en) | Content-based matching of videos using local spatio-temporal fingerprints | |
CN1479910A (en) | Signal processing method and equipment | |
CN1711556A (en) | A method of and system for detecting uniform color segments | |
JP2006092559A (en) | Method of representing at least one image and image group, representations of image or image group, method of comparing image and/or image group, method of encoding image or image group, method of decoding image or image sequence, use of encoded data, apparatus of representing image or image group, apparatus of comparing image and/or image group, computer program, system and computer readable storage medium | |
JP2011108262A (en) | Method for representing at least one image and image group, method for comparing image and/or image group, method for encoding image or image group, method for decoding image or image sequence, usage of encoded data, device, computer program, system, and computer readable storage medium | |
CN1909677A (en) | Image representation and analysis method | |
CN113435391B (en) | Method and device for identifying infringement video | |
US20030026476A1 (en) | Scaled image generating apparatus and method, image feature calculating apparatus and method, computer programs therefor, and image data structure | |
CN1253822C (en) | K near neighbour method used for video frequency fragment fast liklihood inquiry | |
Ventura Royo | Image-based query by example using mpeg-7 visual descriptors | |
Abdelali et al. | A study of the color-structure descriptor for shot boundary detection | |
Ranathunga et al. | Reduction of syntactic video data clustering complexity in processing with compacted dither coding | |
Zargari et al. | A robust compressed domain feature vector for texture based image retrieval | |
Ranathunga et al. | Analysis of video content in multi codec formats with compacted dither coding | |
KR101377460B1 (en) | System for evaluating and certifying roi-based image identify technology | |
Meyer et al. | A novel method for efficient archiving and retrieval of biomedical images using MPEG-7 | |
Bodin et al. | Automated Intelligence Gathering Through Comparison of JPEG Images and their Thumbnails | |
Izquierdo | A low access latency video portal | |
CN1714574A (en) | Non-linear quantization and similarity matching methods for retrieving video sequence having a set of image frames |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C14 | Grant of patent or utility model | ||
GR01 | Patent grant | ||
TR01 | Transfer of patent right |
Effective date of registration: 20190605 Address after: Tokyo, Japan Patentee after: Rakuten, Inc. Address before: Tokyo, Japan Patentee before: MITSUBISHI ELECTRIC Corp. |
|
TR01 | Transfer of patent right | ||
CP01 | Change in the name or title of a patent holder |
Address after: Tokyo, Japan Patentee after: Lotte Group Co.,Ltd. Address before: Tokyo, Japan Patentee before: Rakuten, Inc. |
|
CP01 | Change in the name or title of a patent holder | ||
CX01 | Expiry of patent term |
Granted publication date: 20100623 |